Hi Guys,

I got weird Z11 parameter with SP analysis of spectre simulator. That is a 50 Ω microstrip. So I simulated with a n2port and read the S-parameter file.

I got weird Z11 if I put two port cells at both ports. But Z11 looks good if the port cell at the second port is replaced with a 50 Ω resistor.

The schematics and the Z11 vs. frequency are shown below.

I read Kundert's The Designer's Guide to Spice & Spectre. For SP analysis, only the AC magnitude at one port is set to non-zero at any time. So I think both configuration should give correct result. Who can explain the weird Z11?

Best Regards,
Yawei

The schematic of one port cell on the first port and one 50 Ω resistor on the second port.

Furthermore, for the schematic two_port_cell, Z11 looks good if only one port is selected for SP analysis.
Code:

NPORT0 (PORT0_PLUS 0 NPORT0_t2 0) nport thermalnoise=yes \
	  file="~/test.s2p"
PORT3 (NPORT0_t2 0) port r=50 type=sine
PORT0 (PORT0_PLUS 0) port r=50 type=sine
...
sp sp ports=[PORT0] start=1 stop=2G dec=10 annotate=status  

But Z11 is weird if both port cells are selected for SP analysis.
Code:

NPORT0 (PORT0_PLUS 0 NPORT0_t2 0) nport thermalnoise=yes \
	  file="~/test.s2p"
PORT3 (NPORT0_t2 0) port r=50 type=sine
PORT0 (PORT0_PLUS 0) port r=50 type=sine
...
sp sp ports=[PORT0 PORT2] start=1 stop=2G dec=10 annotate=status 

Thanks for any comments.
Yawei